Volume 141, Issue 5, Pages 1852-1863.e3 (November 2011) Cathelicidin Signaling via the Toll-like Receptor Protects Against Colitis in Mice  Hon Wai Koon, David Q. Shih, Jeremy Chen, Kyriaki Bakirtzi, Tressia C. Hing, Ivy Law, Samantha Ho, Ryan Ichikawa, Dezheng Zhao, Hua Xu, Richard Gallo, Paul Dempsey, Genhong Cheng, Stephan R. Targan, Charalabos Pothoulakis  Gastroenterology  Volume 141, Issue 5, Pages 1852-1863.e3 (November 2011) DOI: 10.1053/j.gastro.2011.06.079 Copyright © 2011 AGA Institute Terms and Conditions

Figure 1 Increased mCRAMP expression in colonic macrophages of mice with DSS-induced colitis. Mice were provided with either water alone or water containing 5% DSS. Animals were killed, and colonic tissues were removed on days 0, 3, and 5. (A) Colonic mCRAMP protein was measured by ELISA. The mCRAMP level was significantly increased in DSS-treated mice when compared with day 0. (B) mCRAMP protein in colons was detected by Western blot analyses. (C) Densitometry of mCRAMP Western blot analyses. (D) Immunofluorescence staining of mCRAMP (red) and F4/80+ macrophages (green), counterstained by nuclear 4′,6-diamidino-2-phenylindole (DAPI) (blue) label. Overlapping expression of mCRAMP and F4/80 in merged images is indicated by a yellowish color at 100× to 200× original magnification. Results are representative of 4 mice per group. (E) Immunofluorescence staining of LL-37 (red) and EMR1+ macrophages (green), counterstained by nuclear DAPI (blue) label. Overlapping expression of LL-37 and EMR1 in merged images is indicated by a yellowish color at 100× to 200× original magnification. Results are representative of 4 patients with UC and 4 healthy patients. Increased colonic cathelicidin protein expression is localized at macrophages and epithelium during colitis. Gastroenterology 2011 141, 1852-1863.e3DOI: (10.1053/j.gastro.2011.06.079) Copyright © 2011 AGA Institute Terms and Conditions

Figure 2 Increased bacterial invasion into inflamed colons of DSS-exposed mice. (A) E coli bacterial cells were cultured and diluted to different CFU per milliliter. The bacterial DNA was extracted and detected by (A) absolute bacterial-specific 16S rDNA real-time PCR and (B) spectrophotometry. Both methods used in A and B show similar correlation between CFU per milliliter and bacterial DNA concentration. (C) PCR band at 500 base pairs showing the presence of bacterial 16S rDNA in colonic DNA samples with positive and negative control groups. (D) Quantitative real-time PCR of 16S rDNA. Bacterial 16S DNA level was significantly increased in the DSS group when compared with the normal group. Using the standard curves in A and B, the approximate CFU per 20 mg colon tissue was calculated. (E) Gram staining for gram-positive bacteria (purple) and gram-negative bacteria (orange) in the colon tissues of healthy and DSS-exposed mice at 200× original magnification. Results are representative of 4 mice per group. Bacterial invasion into colon tissues is evident in DSS-induced colitis. Gastroenterology 2011 141, 1852-1863.e3DOI: (10.1053/j.gastro.2011.06.079) Copyright © 2011 AGA Institute Terms and Conditions

Figure 3 Bacterial DNA stimulates LL-37 expression in human monocytes. (A) Human primary monocytes were incubated with 500 ng/mL of E coli genomic DNA for 0 to 24 hours. LL-37 in conditioned media and cell lysates were measured by ELISA. LL-37 levels in conditioned media were significantly increased at 2 to 24 hours when compared with 0 hours. (B) LL-37 mRNA was determined by real-time RT-PCR. LL-37 mRNA expression was significantly increased at 2 hours when compared to respective 0 hours. (C) Human primary monocytes were treated with 0 to 1000 ng/mL of E coli DNA, and LL-37 mRNA expression was measured by real-time RT-PCR. LL-37 mRNA expression was significantly increased when compared with control. (D) Human primary monocytes were treated with 500 ng/mL of E coli DNA for 0 to 60 minutes. Phosphorylated ERK1/2 and β-actin were determined by Western blot analyses. (E and F) Human primary monocytes were pretreated with dimethyl sulfoxide (control vehicle), ERK1/2 inhibitor PD98059 (10 μmol/L), the nuclear factor κB inhibitor caffeic acid phenethyl ester (1 μmol/L), TLR9 receptor antagonist ODN-TTAGGG (25 μmol/L), the protein synthesis inhibitor cycloheximide (1 μmol/L), and the RNA transcription inhibitor actinomycin D (1 μmol/L) for 30 minutes before incubation with E coli DNA (500 ng/mL) for 4 hours. LL-37 levels in cell lysates were measured by ELISA. The decrease was statistically significant when compared with dimethyl sulfoxide and the E coli DNA–treated group. Results are representative of 3 separate experiments. E coli DNA induces LL-37 de novo mRNA and protein synthesis in monocytes that are ERK and TLR9 dependent. Gastroenterology 2011 141, 1852-1863.e3DOI: (10.1053/j.gastro.2011.06.079) Copyright © 2011 AGA Institute Terms and Conditions

Figure 4 E coli DNA induces TLR9-dependent LL-37 expression in human monocytes. (A) Monocytes were cotransfected with control, TLR9, and MyD88 siRNAs. Transfected cells were treated with E coli DNA (500 ng/mL) for 4 hours. Cell lysates were used for LL-37 ELISA. The decrease was statistically significant when compared with the control siRNA transfected E coli DNA–treated group. (B) Western blot showing the successful knockdown of TLR9 and MyD88 by respective siRNA. (C) Experimental plan of multiple intracolonic E coli DNA administrations to healthy and DSS-exposed mice. WT or mCRAMP-deficient mice (∼20 g) were given 5% DSS in their drinking water or water alone and injected intracolonically with 2.5 mg/kg of E coli DNA in 50 μL. (D and E) Colonic levels of mCRAMP (D) protein and (E) mRNA on day 5 were measured by ELISA and real-time RT-PCR. Colonic mCRAMP expression is increased after DSS treatment and is further augmented by E coli DNA administration. (F) Colonic levels of TLR9 mRNA. TLR9 mRNA expression was significantly increased after DSS treatment but was reduced by intracolonic E coli DNA administration. Each group includes 6 mice. Gastroenterology 2011 141, 1852-1863.e3DOI: (10.1053/j.gastro.2011.06.079) Copyright © 2011 AGA Institute Terms and Conditions

Figure 5 TLR9 mediates colonic cathelicidin expression in vivo. (A and B) TLR9-deficient mice and WT mice were given 5% DSS in their drinking water or water alone for 5 days. Colonic levels of mCRAMP (A) peptide and (B) mRNA were measured by ELISA and real-time RT-PCR. The decrease was statistically significant when compared with DSS 5% treated WT mice. Each group includes 4 mice. After exposure to DSS, colonic mCRAMP levels are lower in TLR9 KO than those of WT mice. (C) Histologic score of TLR9 KO and WT mice. TLR9 KO mice developed significantly more severe histologic damage than WT mice after DSS exposure. Gastroenterology 2011 141, 1852-1863.e3DOI: (10.1053/j.gastro.2011.06.079) Copyright © 2011 AGA Institute Terms and Conditions

Figure 6 Endogenous mCRAMP modulates course of DSS-induced colitis. WT and mCRAMP KO mice (∼20 g) were treated with 2.5 mg/kg E coli DNA intracolonically in 50 μL and provided with water containing a 5% DSS solution for 5 days. (A and B) Body weight change of mice. (A) DSS exposure led to a significant decrease of body weight, which was augmented in mCRAMP KO mice and further worsened by intracolonic E coli DNA administration. (B) No significant change in body weight of all water-treated groups. (C) H&E staining of colons of mice. (D) Clinical score and (E) histologic score of DSS-induced colitis of various groups of WT and mCRAMP KO mice. The difference between control WT mice and mCRAMP KO mice was statistically significant. Also, the increase of clinical score and histologic score in the E coli DNA–treated mCRAMP KO group was significant when compared with the control mCRAMP KO group. (F) TUNEL staining of colons of DSS-exposed mice with apoptosis index. Apoptotic cells were visualized as brown spots. mCRAMP KO mice generally develop more serious colonic apoptosis than WT mice after exposure to DSS. Intracolonic administration of E coli DNA exacerbates apoptosis in mCRAMP KO mice but not in WT mice. Results are representative of 4 mice per group. Gastroenterology 2011 141, 1852-1863.e3DOI: (10.1053/j.gastro.2011.06.079) Copyright © 2011 AGA Institute Terms and Conditions

Figure 7 Bone marrow–derived cathelicidin modulates colitis. (A) Plan of bone marrow transplant experiment with clinical score of DSS-induced colitis groups. H2O-treated healthy groups had all a score of 0. (B) Body weight change of mice that underwent a transplant. (C) H&E and TUNEL staining of colonic tissue (original magnification 100×). (D) Histologic score of mice that underwent a transplant. (E) Colonic mCRAMP protein levels of mice that underwent a transplant. (F) Colonic tumor necrosis factor-α mRNA levels of mice that underwent a transplant. WT bone marrow transplant to mCRAMP KO mice increases colonic mCRAMP mRNA levels and ameliorates colitis. KO bone marrow transplant to WT mice decreases colonic mCRAMP mRNA levels and worsens colitis. Results are representative of 6 mice per group. Gastroenterology 2011 141, 1852-1863.e3DOI: (10.1053/j.gastro.2011.06.079) Copyright © 2011 AGA Institute Terms and Conditions

Supplementary Figure 1 Representative figure of flow cytometry data of bone marrow and peripheral blood cells showing the engraftment of bone marrow transplant. Gastroenterology 2011 141, 1852-1863.e3DOI: (10.1053/j.gastro.2011.06.079) Copyright © 2011 AGA Institute Terms and Conditions